Even before 2020’s first horrific wave of COVID-19 deaths subsided, reports surfaced warning of a brutal second punch: Instead of recovering quickly after a mild infection, some people were suffering from symptoms that lingered or even intensified in the weeks and months that followed.
The condition came to be called long COVID. In those early days, everything about it was uncertain, from what symptoms it caused to how long they’d last and how hard they would hit. Some speculated that the effects might be effectively incurable, and that a large percentage of those infected with SARS-CoV-2 would wind up succumbing to this life-altering condition. “Months of illness could turn into years of disability,” warned the Atlantic’s Ed Yong. Given the bodies piling up in makeshift morgues, it seemed reasonable to assume the worst.
The first reports of how COVID seemed to fundamentally change people were scrabbled together from anecdotal accounts and preliminary studies. The picture they painted was frightening: As many as a third of all people who’d tested positive went on to report long COVID, according to a report published a year after the pandemic began. Few of these people had recovered much, and many were debilitated, unable to work or attend school. Newspapers and magazines ran articles that vividly described the complex litany of suffering endured by patients who weren’t getting any answers from their doctors. One early and influential story was by the British epidemiologist Paul Garner, who wrote in the medical journal BMJ about being flattened by a “roller coaster of ill health, extreme emotions, and utter exhaustion.” He described experiencing relentless, extreme fatigue, a “muggy head,” breathlessness, muscle pain, and a “weird sensation in the skin”—a parade of “constantly shifting, bizarre symptoms” that left him bedridden.
Long COVID is an unusual condition not only in its kaleidoscope of symptoms but also in the fact that it hadn’t been identified initially by doctors who encountered similar sets of symptoms in their patients. It was, rather, described by COVID patients themselves who, in the early months of the pandemic, found themselves mysteriously unable to get better. The complaints of early “long-haulers” were then picked up and amplified by activists, whose lobbying persuaded the government to allocate more than $1 billion in research funds. “Long COVID has a strong claim to be the first illness created through patients finding one another on Twitter,” researchers Felicity Callard and Elisa Perego wrote in the journal Social Science & Medicine. (They both suffered from long COVID themselves.) Patients desperately searching for answers were understandably dismayed to find little clarity from the medical community about their strange illness.
Now, three years later, the research is catching up to the anecdotal reports and the early evidence, and a clearer picture of long COVID has emerged. It turns out that, like COVID-19 itself, a lot of our early guesses about it turned out to be considerably wide of the mark. This time, fortunately, the surprises are mostly on the positive side. Long COVID is neither as common nor as severe as initially feared. As the U.S. government moves to end the country’s state of emergency, it’s another reassuring sign that, as President Biden put it during his State of the Union address, “COVID no longer controls our lives.”
As vaccines rolled out across the country in 2021, researchers at the Mayo Clinic analyzed the symptoms of 108 patients who’d come for post-COVID care. Their results suggested that these patients fell into two main camps. Some, mostly men, suffered severe illness and were still being plagued by symptoms like chest pain and shortness of breath. Then there were others, mostly women, who had experienced relatively mild illness, or no symptoms at all, but were subsequently dogged by “widespread pain, fatigue,” and “cognitive impairment, including the commonly reported ‘brain fog.’ ” The authors noted that this cluster of symptoms resembled a class of broadly similar conditions like chronic fatigue syndrome, fibromyalgia, and POTS (postural orthostatic tachycardia syndrome), all of which can leave sufferers incapacitated for years at a time.
How long could long COVID symptoms be expected to last? Researchers in Australia tried to answer that question by conducting phone interviews with every single person who was diagnosed with COVID-19 in the state of New South Wales between January and May 2020. They found that recovery followed a curve, with 80 percent of patients fully recovered after 30 days and 91 percent recovered after 60 days. Thereafter, the population of symptomatic patients continued to slowly dwindle, with 4 percent of the original patient population still suffering symptoms after four months. Their most common complaints were coughing and fatigue.
Other work suggested that long COVID could affect a much larger slice of the population. In one influential study from early 2021, researchers at the University of Washington sent a questionnaire to 234 COVID patients between three and nine months after they fell ill. Of the 177 who responded, about a third reported ongoing symptoms like fatigue, brain fog, and loss of smell. A subsequent Brookings Institution report used this statistic to estimate that 31 million working-age Americans “may have experienced, or be experiencing, lingering COVID-19 symptoms.”
There are several problems with survey-based research, however. One is that there’s a risk of selection bias, in that people who feel that they have long COVID are more likely to want to complete a questionnaire on the topic. Another is that people may report having symptoms post-COVID that they also had pre-COVID, and so their maladies may not actually have been caused by the disease.
To get around these problems, scientists began carrying out what are called retrospective cohort studies. These involve sifting through anonymized electronic medical records to find patients who tested positive for COVID and then returned complaining of subsequent symptoms. Patients who experienced the same symptoms both before and after they got COVID are filtered out. Those remaining are then compared with a second population, of COVID-negative patients, with whom they have been matched by age, gender, and other medically relevant criteria. The difference in the groups’ rate of post-COVID symptoms reveals just what medical mayhem the SARS-CoV-2 virus is leaving in its wake.
This kind of research isn’t quick, because, by definition, it concerns the patients whose maladies take the longest to resolve. But as 2022 progressed, results started to come in.
One study of patients in an Israeli health network looked at the incidence of 70 commonly reported long COVID symptoms in 150,000 patients. The researchers found that patients who’d been infected were more likely than people in a control group to suffer for extended periods from certain symptoms, in particular loss of taste and smell, concentration and memory problems, difficulty breathing, weakness, hair loss, palpitations, and chest pain. But the difference between the infected and controls largely disappeared by the end of the first year, and to the extent that they remained, they affected a relatively small number of patients. For instance, 407 of the COVID patients reported having persistent concentration and memory problems at the end of the first year, while 276 of the controls also did. That meant that for every 10,000 people, only about 13 had cognitive difficulties that were attributable specifically to COVID.
The researchers had gone into the project expecting to find a large number of chronic COVID aftereffects. Instead, they concluded that there were actually very few. “As we analyzed the data,” the lead authors told Stat in January, “we were surprised to find only a small number of symptoms that were related to COVID and remained for a year post infection and the low number of people affected by them.”
Other studies produced similar results. Researchers at the University of Oxford in the U.K. combed through the health records of more than a million patients in a retrospective cohort study that compared those who’d tested positive for COVID with those who’d had other respiratory infections but had not been diagnosed with COVID-19 or tested positive for SARS-CoV-2. After following patients’ symptoms for two years, they reported in the Lancet Psychiatry last August that they “found no evidence of a greater overall risk of any first neurological or psychiatric diagnosis after COVID-19 than after any other respiratory infection.” There was an elevated risk for certain symptoms, however. They found that 6.4 percent of COVID patients experienced “cognitive deficit (known as brain fog),” compared with 5.5 percent of patients who’d had other respiratory infections. Although the Oxford researchers were looking at a different set of cognitive symptoms than the Israeli researchers were, the upshot was similar: In both cases, nearly as many controls suffered the symptom as COVID patients did.
Meanwhile, researchers at Montefiore Medical Center in the Bronx looked at 18,811 patients who’d tested positive for COVID-19 and 5,772 who’d had influenza. The number of patients reporting new-onset neuropsychiatric symptoms after COVID-19 was 388, or 2 percent. This figure was actually less than that for patients with influenza, which was 2.5 percent.
There’s another way to look at long COVID’s impact, and that’s by examining how it has affected the workforce. “The COVID-19 pandemic will almost certainly create a substantial wave of chronically disabled people,” Ed Yong wrote in 2020. Others argued that this surge of long-haul cases would not only mean enormous suffering but would actually pose a threat to economic recovery. “Long COVID is contributing to record high numbers of unfilled jobs by keeping millions of people from getting back to work,” a Brookings report suggested last year.
There is no evidence that any of this has actually happened. Not only did disability claims not rise during the pandemic, they fell. “You see absolutely no reaction at all to the COVID crisis,” Nicole Maestas, an associate professor of health care policy at Harvard, told Benjamin Mazer of the Atlantic in June 2022. “It’s just not a mass disabling event.”
Further data bear this out. In January, the New York State Insurance Fund, which administers disability claims, released a report analyzing long COVID claims made between Jan. 1, 2020, and March 31, 2022. It found that while there were several hundred successful claims after the initial wave in March and April of 2020, the number subsequently fell to fewer than 10 per month, and spiked into the double digits only after the alpha and omicron waves. “The percentage of people meeting the report’s definition of long COVID in the overall COVID claimant population is declining,” said Gaurav Vasisht, the NYSIF’s CEO and executive director. The most recent data, from March 2022, shows that only about 5 claims for long COVID were being granted per month out of about 3,000 disability claims in the entire state.
The best available figures, then, suggest two things: first, that a significant number of patients do experience significant and potentially burdensome symptoms for several months after a SARS-CoV-2 infection, most of which resolve in less than a year; and second, that a very small percentage experience symptoms that last longer. I want to be clear about this: Long COVID is a real illness that has dramatically affected many people’s lives. But its prevalence does seem significantly less worrisome than originally thought.
Another insight that emerges from the cohort studies into long COVID is that it’s not so easy to prove causality between a particular infection and a symptom. Almost all the symptoms associated with long COVID can also be triggered by all sorts of things, from other viruses to even the basic reality of living through a pandemic. Fatigue, for instance, can be caused by COVID-19—or by stress, depression, sleep disorders, anemia, and cancer, among other things. So, even though many patients insist that they are COVID long-haulers—and their symptoms align entirely with the common understanding of the condition—it’s entirely possible that they’re dealing with something slightly different from long COVID. Data from the Census Bureau and the National Center for Health Statistics released in January, collected in 20-minute online surveys, shows that 11 percent of American adults who have had COVID say they are currently experiencing lingering symptoms. But this self-reported information has not been borne out by more rigorously collected data. In the absence of any test for the disease, there is no way to definitively say that their symptoms are actually directly caused by the SARS-CoV-2 virus.
There is no question that many people experiencing long COVID (or something like it) are struggling, both with symptoms and with a medical community that often fails to properly treat them. There are patients describing debilitating fatigue and neuropsychiatric symptoms today, just as there were in 2020; they are still expressing a sense of frustration at the lack of answers they’re getting from doctors. In February, Atlantic writer Katherine Wu described a Brazilian long COVID patient whose ordeal sounds worryingly similar to those of the very first long-haulers: Her “days revolve around medications and behavioral modifications she uses for her fatigue, sleeplessness, and chronic pain,” and she “no longer has the capacity to cook or frequently venture outside.” The patient tells Wu: “Sometimes I think the person I used to be died in April of 2020.” Even if the cohort of these patients with long COVID represents a very small percentage of those who have been infected with SARS-CoV-2—and even if it includes people suffering not from COVID but from other lingering viruses or underlying conditions—their needs are real.
But one cannot conclude that there is, as Wu puts it, “an ocean of patients with titanic needs” attributable specifically to COVID. The ocean of patients who experience long COVID appear to do so usually only to a moderate degree and for a limited time; the titanic needs are experienced by a relative few.
Indeed, even patients who suffer the most debilitating form of long COVID can find themselves improving after weeks and months, not years. In January 2021, eight months after his first essay ran in the BMJ, Paul Garner published a follow-up article in which he described how he had achieved a total recovery through a process that included plenty of exercise. “I’m back to normal,” he told me. Today he believes that the perception of a long COVID public health crisis—one his own experience helped to fuel—is overblown. “The predicted long-term disability from tissue damage simply has not materialized,” he said. “The public narrative has morphed into continued catastrophic thought.”
Despite the ongoing failure of the long COVID tsunami to arrive, the media has continued to sound the alarm. Last month the editorial board of the Washington Post ran an opinion, under the headline “Long covid haunts millions of people,” that is typical of the prevailing mindset. The piece contends that while “it is not yet known how many people have long COVID, why and what their prospects for recovery are, let alone what the long-term impact on society will be,” the condition nevertheless must be regarded as a serious problem: “The entire world will have to prepare for a legacy of long-COVID sufferers.”
That’s pretty much the same attitude Yong took when he wrote about long COVID in 2020. But back then, there was no data; everything was based on anecdote and speculation. Now COVID should have a clearer foundation. But it doesn’t. Defining long COVID “isn’t an easy thing to do,” said Paul Glasziou, a professor of medicine at Bond University and director of the Institute for Evidence-Based Healthcare in Queensland, Australia, “because there is no absolutely clear boundary that you can draw around the condition.” Patients can present with dozens of symptoms in any variety of combination, and none of them are unique to the ailment. To be blunt, long COVID doesn’t much look like what would normally be a called a disease.
As Science columnist Derek Lowe wrote last year, “So far there are no diagnostic findings that would allow you to even say for sure that post-Covid even exists, biochemically.” (Critically, this means that no test can definitively diagnose long COVID as a unique condition. Studies have shown that symptoms following a COVID infection are associated with real changes in the body.)
Instead, it looks more as if people who complain of long COVID are suffering from a collection of different effects. “I think there’s quite a heterogeneous group of people all sailing under the one flag,” said Alan Carson, a neuropsychiatrist at the University of Edinburgh in Scotland. Some patients may be experiencing the lingering aftereffects that occur in the wake of many diseases; some patients with chronic comorbidities might be experiencing the onset of new symptoms or the continuation of old ones; others might be affected by the sorts of mood disorders and psychiatric symptoms you’d expect to find in a population undergoing the stress of a global pandemic.
“I think the problem is, this was given a name before anyone had really worked out what it was,” Carson said. “And from then, people have been playing catch-up and trying to work out what the thing is, rather than the more traditional way of trying to work out what things are and then naming them accordingly.”
What’s important to understand, said Dacre Knight, a professor of medicine at the Mayo Clinic in Jacksonville, Florida, is that whatever symptoms they experience, most long COVID patients do get better. “There’s a lot of talk about long COVID that has made patients concerned that, because they’ve had symptoms for more than a week or two, maybe they’re going to be stuck with it for months and years,” he said. “So it’s reassuring for them to know that the majority actually do recover with full resolution of their symptoms.”
Now that cases of COVID have started to recede, there’s evidence that long COVID is fading too. The Census Bureau’s survey data—however imperfect a measurement—show that the percentage of COVID-19 patients who currently say they have long COVID fell by 42 percent between June 2022 and January 2023. Knight said he’s seen a decline in the number of long COVID patients coming into his clinic. “It’s just gradually decreasing,” he said. Last year, his clinic was typically seeing 15 to 20 new patients a month; now, he said, “I would say maybe 10 to 15.”
For those whose fatigue, sleeplessness, and chronic pain have proven more stubbornly persistent and debilitating, the tremendous concern about long COVID over the past few years might ultimately prove to be a blessing, regardless of whether it’s fair to call long COVID itself a major public health crisis. Whether their diagnosis is for chronic fatigue syndrome or fibromyalgia or POTS, many other people have been suffering such symptoms for a long time too, and long COVID research has demonstrated that viruses are capable of tinkering with the body in ways we do not yet understand. If research funding unleashed by fears of long COVID ultimately explains why, then that itself would be a significant achievement—an acceleration toward the day when not only long COVID patients but a lot of other suffering people as well will be able to put their illnesses behind them for good.
Update, March 20, 2023: This article has been updated to clarify that research has identified biological changes that can occur in association with symptoms present following a COVID infection.